Last week I had the opportunity to attend an Internet Innovation Alliance Symposium at the Waldorf-Astoria called "The Exaflood - Finding Solutions."
It consisted of a pair of panels stocked with industry luminaries discussing the reality and challenges of the exaflood. Going into the event I had some trepidation that it would be a one-sided defense of the exaflood concept along with a steady drumbeat against legislation, but that couldn't have been further from the truth.
Instead what I got was a lively, in-depth discussion that covered a number of interesting points, so many, in fact, that I've decided to split my coverage of the event into two parts.
For part one I'll dive into the first panel, which featured: Paul Mankiewich, CTO of Alcatel-Lucent; Andrew Odlyzko, who researched trends in Internet demand at the University of Minnesota; Nick Rockwell, CTO for MTV Networks; and Bret Swanson, senior fellow at the Progress & Freedom Foundation and director of the Center for Global Innovation.
I was most intrigued by the involvement of Andrew Odlyzko. He's been a leading voice tempering worries about the exaflood as his research has shown that while for the first decade Internet traffic was doubling year over year, that recently it's only been increasing by about 50%.
In his remarks and subsequent discussion, he alluded to the historical context that people were worried about the petaflood a few years ago but nothing bad happened. He even went so far as to say that with a 50% growth rate that he doesn't see the need for giant new investments in capacity.
At the same time, his perspective was quite nuanced. On the one hand while citing the projections network operators have made for continued growth in demand, he lamented that if Qwest's projections of 19% growth that that would likely be disastrous for the continued adoption of the Internet. On the other he pointed to the fact that while South Korea has only one sixth the population of America that they generate the same amount of demand for bandwidth, and that it'll take 5 years for us to catch up to where they are today, let alone where they'll be in the future.
Where Andrew sees the biggest challenges is in his observation that the real money isn't in delivery high bandwidth to consumers. And, in fact, he believes network operators are facing a bit of a dilemma: should they be limiting or encouraging growth in demand for bandwidth? By maintaining the status quo he feels they can support the current 50% growth rate without need for major investment, but if they try encouraging growth that could push the rate to 100% year over year and therefore demand greater investment.
These ideas get to the heart of one of the bigger problems we face in getting more capacity deployed. Broadband providers have trouble making more money off of faster service, and because of the all-you-can-eat model prevalent in America the more the customers use their networks the tighter their margins get. I worry that what this all leads to is a system where the best scenario for network operators is 100% adoption and retention but 0% use and only enough investment to keep people placated.
The comments of Nick Rockwell added the interesting dimension of a content owner's perspective into the mix.
One thought he shared is that technologies like the Slingbox, which are driving significant demand for upstream bandwidth, are actually workarounds trying to overcome the limitations of the current distribution system. He even went so far as to say that in a perfect world the Slingbox wouldn't be necessary as that functionality would already be in place.
Nick gets credit for having the best line of the event. Playing off of the common theme of "if we build it they will come" he said, "If we don't build it they can't come." Continuing on he argued that there needs to be excessive capacity in order to drive prices down and innovation up. In his mind, no amount of bandwidth is enough, and as an innovator he'd love it if bandwidth capacity ran ahead of his ability to make use of it, which hasn't been the case anywhere until recently.
He then took that idea to a global level, suggesting that innovators are developing products now in South Korea that you just can't do in the US today with our current broadband infrastructure. And beyond that, until we have that level of infrastructure, innovators can't really even be thinking about what they could do with all that capacity as they have to focus on what's possible with what they have to work with today.
This was a sentiment echoed by Brett Swanson. He firmly stated that abundance must come first, that abundance leads to innovation, and that by deploying what seems like too much now we will find new innovations.
I asked the panel how much bandwidth is enough and Brett was the only one to give a specific answer: 50-100Mbps in the next 5-10 years. He then also alluded to the fact that cable companies can reach this level and beyond as they start looking at moving to a switched network.
While none of this was necessarily new info to me, it was kind of startling to have these thoughts all presented at once as what it looks like people are saying is that we're already at a competitive disadvantage to countries like South Korea when it comes to enabling an environment that fosters innovation. And not only that, because of the time and effort it takes to deploy new networks, every day that passes that we're not moving forward aggressively to catch up we're falling further behind.
To put it more succinctly: we're already five years behind in positioning our country to be leaders in the 21st century, and until we get our act together we're just going to keep falling further back as our position as a global Internet leader continues to be threatened.
Throughout this panel I found the contributions of Paul Mankiewich to be fascinating. He's a guy who's literally helping build the Internet, going so far as to refer to himself as a plumber rather than a tech guy.
While I've often heard that the Internet wasn't designed to support all the interactive rich media applications that have sprouted up I've often disregarded that sentiment. But it's a different matter entirely when the CTO of a company as significant as Alcatel-Lucent bluntly says that they didn't predict the explosion in content created by users and that the idea that people are pushing up content is something they're still struggling with.
This isn't giant network operators trying to claim the Internet isn't ready for this as a cover for them to deliver less service or degrade the service they already provide, this is one of the guys who designs the underlying technology of the Internet saying they weren't expecting this remarkable transition of the Internet beginning to realize its full potential as an interactive medium.
Now, to some degree I'm really disappointed to hear this. Why couldn't people have predicted that there'd be demand upstream as well as downstream? Hasn't the Internet always been touted as a two-way communications tool?
But at the same time I understand that the dominant paradigm of all mass media up until the Internet has been consumer as passive recipient rather than active user.
Even still, Paul helped point out that many of the challenges the Internet faces today aren't technological limitations but instead business processes like DRM that are preventing content from being stored in the network.
The final thought I'll leave you with today came from Nick, who's response to questions about the need for legislation to guide network management decisions was that we need to listen to the engineers. He believes that they've done a great job scaling the Internet to date, and that we should let the people who built the networks solve the problems.
While he didn't explicitly say so, I'm guessing that he includes in this also the input of engineers who make the applications that run on the networks.
I think too often we're allowing debates around issues like net neutrality to be driven by public interest groups that aren't taking into consideration the intricacies of network deployment and management. And on the flip side, the network operators aren't being open enough about the challenges they face.
If we could just get all these great engineering minds in the same room, from both the application and network sides, I feel like it wouldn't be all that difficult to find solutions to these problems.
Of course, the challenge is that we can't ignore the business side of these issues, and that many of the decisions being made about how to run networks technologically will have a huge impact on the businesses of everyone involved.
The gist of this first panel seemed to be that while yes there is tremendous growth in demand for bandwidth that it is not yet a significant problem in terms keeping networks up and running but that we are facing tremendous challenges if we want to remain global leaders in Internet innovation.
More to come tomorrow from the second panel!